fix wrong float_t on x86_64

x86_64 does not have excess precision, at all
1 file changed